Re: Group selection in the breeding of super chickens
- From: "John W Edser" <edser@xxxxxxxxxxxxxx>
- Date: Sun, 10 Feb 2008 01:17:27 -0500 (EST)
Lorentz <drosen0000@xxxxxxxxx> wrote:-
JE:-
You are not new to this group. I must have defined these over a hundred
times by now. ...
1) Total Darwinian Fitness (TDF): the total number of just fertile forms
reproduced per parent per population. The only possible way to halt all
Darwinian natural selection within one natural population is to
artificially
maintain TDF to remain equal per parent per population. Note that this
also
remains the only way that heritable variation provided by just random
mutation and random sampling error can become empirically separated
within a
natural population because after non random selection has been halted
entirely, all that remains are these random processes.
Your formulation STILL doesn't explain how a sexual animal can go
from being almost totally asocial to have a refined, sophisticated
group identification.
JE;-
Sexual forms can indeed be selected to become social in more and more
complex ways via mutualised increases in TDF. Groups are predicted to form
and split depending on mutualised TDF payoffs. The mean cost of maintaining
such groups (they don't come for free) must remain less than the mean TDF
mutualised gains otherwise the social grouping will be selected against at
just the monocentric Darwinian, adult organism level of selection and
dissolve. The mutual reduction of risk per parent has become a very strong
focus for the selection of social group mutualisation via the Baldwin
effect. An example is the way Emperor penguins huddle to provide warmth when
incubating an egg with their feet in just appalling, sub zero conditions.
Cheating can _lower_ the mean TDF mutualisation point providing a loss for
every parent within one group _including the cheater_.
http://en.wikipedia.org/wiki/Emperor_Penguin
Those huddling on the outside are more at risk than those on the inside
where the larger the group the less are the relative numbers that have to be
on the outside because of simple volume/surface area ratios. Mutualised TDF
selection predicts that these groups should remain as large as possible.
Study of these huddles demonstrates that every penguin must rotate their
position so that those on the outside eventually get their turn on the
inside. Note that any one penguin not rotating stops all the others. An
insurance premium which has to be paid (your turn on the outside) provides
a gain larger gain (the time you spend on the inside) _for each and every
member of that group_ allowing a complex social trait to evolve without any
group selection. Note that some social interrelation is much better than
none at all, i.e. just two penguins making a huddle is much better than
none. Primitive, selfish, scramble competition for huddling which acts as
a critical _social resource_ cannot work because the ensuing fighting
destroys that resource. This is just obvious in this unique situation.
However, I argue that the exactly the same risks/costs are to be found
within _any_ fitness mutualized social trait. The only difference is that
they are not as obvious. In the chicken example the socialised group paid
monocentric TDF dividends which remained greater than the costs allowing
group sociality to remain 100% NOT group selected.
The self contradiction can be seen in three easy
steps.
1) Your formulation doesn't allow us to consider the natural selection
of a gene. Talking about competition between genes is a narrow, gene
centric view. I can't talk about selfish genes, altruistic genes,
cheater genes, or even genes that make one good looking. I can only
talk about genes in the "sexually mature individual" or some
approximation of that concept.
JE:-
Monocentric selection acting on single genes can happen, but only indirectly
and not directly. This is because the fitness relationship of any one gene
to any other within the same genome was and remains epistatic, i.e. non
additive. This can provide two _opposing_ effects: buffer one phenotype
against change (canalization) while providing the reverse for another
phenotype coded within the same genome: accelerate change (assimilation).
Geometric change is not as predictable as simple, additive change, i.e. it
remains many times more complex and not at all understood. This is why it is
deleted within popular gene centric oversimplified models. Please refer to
the Waddington's critical amendment to Haldane's basic population genetics
equations which I have posted here twice, where Waddington provided a
variable for epistasis for the very first time. Waddington's critical
amendment remains ignored.
2) The "sexually mature individual" never replicates itself
completely.
JE:-
Yes. If it did and this was heritable, then evolution by natural selection
would become impossible.
Each sexually mature reproductively active individual
animal (your favourite unit)..
JE:-
No, just Darwin's _implied_ monocentric unit.
will vary greatly from it parents. Since
sexual animals exchange genes (the units you feel are too poorly
defined to be meaningful), each generation is different both
genetically and phenotypically from the next.
JE:-
Yes, but the overwhelming majority of heritable variation is deemed
"inherited but not heritable" only because it remains epistatic.
Conventional wisdom dictates that epistatic traits cannot survive meiosis as
just an unproven proposition.We simply cannot code for the vast number of
inherited phenotypes which typify our species compared to another such as
the chimp with only 20,000 or so additive genes.
We may resemble our
parents, but we never are exactly like our parents.
So we don't know
what variation is acted on by natural selection, because "the whole
animal" doesn't have any trait that is 100% inheritable.
JE:-
And..neither does the _fitness_ of any one gene...
If a "sexually mature individual" is very fit, I can't discuss which of
its
siblings carries the mantle of being its true copy.
JE:-
Canalisation can take care of that.
3) By your definitions, there can be no sexually mature individuals
individuals that can be described as "group of more than one sexually
mature individual organisms."
JE:-
There are expanding and contracting fitness mutualised groups of sexually
mature individuals which associate or not, depending entirely on the
benefits they provide for others /take for themselves, on just an additive
fitness basis.
Otherwise, one is left with the same
ambiguities as in a "gene centric model." So I can't use group
selection in your formalism, either.
JE:-
My formalism of Darwinism allows just the one fitness maximand: TDF. A
maximand is something which _always_ remains maximal e.g. the velocity of
light in a vacuum. If two maximands exist within the same theory they must
mutualise or extinguish each other. Di-centric group selection theory has
never taken this logic into consideration.
You can't make an ambiguity go away by saying "statistics lie."
Regardless of whether a model can be mathematically modeled or not,
the logic has to be describable linguistically.
JE:-
I have never claimed that "statistics lie" because they always have to be
interpreted within a theory and I agree entirely that all valid models
remain logical. I propose that any linguistic model has to remain a theory
and not just a simplified/oversimplified model of a theory
A linguistic ambiguity
is still an ambiguity.
JE:-
I entirely agree.
Its possible there is no completely unambiguous
formulation that completely describes natural selection.
JE:-
No. TDF equalisation remains _the only empirical way_ to halt all Darwinian
monocentric natural selection within one natural population.
Never the
less, maybe you should document the points in your theory, and others
if you wish, where the concept is ambiguous.
JE:-
AFAICT, Darwinian theory as I have formulated it remains unambiguous.
Regards,
John Edser
Independent Researcher
edser@xxxxxxxxxxxxxx
.
- Prev by Date: Re: Group selection in the breeding of super chickens
- Next by Date: Re: Group selection in the breeding of super chickens
- Previous by thread: Re: Group selection in the breeding of super chickens
- Next by thread: Re: Group selection in the breeding of super chickens
- Index(es):
Relevant Pages
|
